Showing papers by "Sara Kiesler published in 1993"

Returns to science: computer networks in oceanography

Abstract: ecent developments in network communication and scientific computation have led to calls for a massive increase in computer network support for science [22]. Computer science has benefitted from networking technology since the Arpanet began in the late 1960s [16]. Other scientific disciplines began using networks in a more or less organized Pfashion by the early 1980s: Highenergy physicists use Hepnet, and molecular biologists use Bionet. Yet network support for science has been raainly an ad hoc enterprise. One vision of the future is the Collaboratory: a \"combination of technology, tools, and infrastructure that allows scientists to work with remote facilities (co-laboratory) and one another (collaboratory) as if they were colocated and effectively interfaced\" [17]. In this article we examine the alternative theories of the returns to science from networking, present early results from one group of networked research scientists, and discuss some implications for policy and technical development. The !general case is familiar: benefits to science from network technology derive from scientists' increased access to remote scientific resources. These resources include scarce and expensive technology such as the space telescope or supercomputers, large databases such as those for DNA sequences or satellite weather images, or scientificliterature databases such as Chemical Abstracts. They include services such as\" mail-order\" fulfillment of requests for reprints, algorithms, or even experimental-chip fabrication. They also include the time and attention of other scientists, accessible through electronic mail, b-boards, and teleconferences. Increasing ease of access to these resources should lead to a concomi tan t increase in scientific productivity. This argument has gone unchallenged even though, generally speaking, computer technologies have not had empirically strong effects on productivity in any industry. In his wideranging review of sectoral, industrial, and organizational studies, Attewell [2] shows that these studies fail to demonstrate a positive association of investments in computer technologies with productivity. For instance, P. Weill [27] divided computer investments of 33 strategic business units into \"strategic systems\" (e.g., an inventory system allowing sales staff to give accura te del ivery t ime estimates), \"transactional systems\" (e.g., accounts payable, order entry), and information systems (e.g., email, networked databases). Transactionalsystem investments (23 % of the total) were related to better performance as measured by improved Return on Assets (ROA) and lower nonproduction labor costs adjusted for sales. But strategic-systems investments (which had a negative effect in the short run) and in fo rmat ion systems were unre la ted to any p e r f o r m a n c e measure. Scientists who use computer networks extensively claim that those networks have already enhanced their productivity and changed the nature of their work [16, 20, 21, 24]. Despite such claims and confident predictions for the future, there has been no systematic empirical research on how computer network support for science is related to important scientific outcomes, including productivity [4, 8, 9, 11]. 1 We therefore decided to do such a study and collected data from